Preclinical Studies Real-time imaging of convection-enhanced delivery (CED). Because the volumetric and anatomic distribution of infusate will differ with treatment site and because various pathologic conditions will cause differences in tissue properties that affect CED parameters, it will be important to monitor CED delivery in real-time to further develop and perfect this delivery method in the clinical setting. To image CED in real-time, we have developed small and large molecular weight computed tomography (CT)- and magnetic resonance (MR)-imaging tracers that can be co-infused with therapeutic agents. We have shown that by combining (or co-infusing) therapeutic molecules and surrogate imaging tracers, CED of putative therapeutic agents can be precisely monitored in real-time using serial CT- or MR-imaging. The capability to non-invasively monitor infusate delivery in real-time permits exploration of a variety of parameters (i.e., rate, effect of flow characteristics, effect of anatomic boundaries) associated with CED, reveals areas for improvement in the CED technology (i.e., catheter design, pump design), improves the infusion accuracy/reliability, confirms adequate target treatment, and permits determining if an infused agent is efficacious if delivered to the target tissue. Preclinical to Clinical Therapeutic Applications Exploiting the unique delivery properties of CED has allowed investigation of new paradigms for the research and treatment of central nervous system (CNS) disorders. We ended a clinical study using IL13-pseudomonas exotoxin to treat diffuse intrinsic pontine glioma (brainstem glioma) because the manufacturer discontinued production of this agent. A manuscript describing the results of this study is under review. We used a bench-to-bedside approach to treat the neurodegenerative disorder, Parkinson's disease, by convective delivery of Adeno-Associated Virus type 2 carrying the Human Glial cell line-Derived Neurotrophic Factor gene (AAV2-hGDNF). The study used escalating doses of AAV2-hGDNF, with 6 patients being treated at the lowest 2 doses and 1 patient being treated at the next highest dose. After treating 13 subjects we stopped enrollment due to slow patient accrual. We will continue to follow all these subjects until 5 years after the last subject was treated. We will report the study results shortly thereafter. Neuro-Oncology. Diffuse infiltrative brainstem gliomas are pediatric brain tumors that are uniformly fatal (median survival of less than 1 year). Complete surgical resection is not possible, and radiation is only palliative. Putative therapeutic compounds have been developed and are available to treat diffuse brainstem gliomas but have not been effective when delivered systemically because they cannot cross the blood-brain barrier into the tumor. To overcome this limitation, we investigated the possibility of using CED of a targeted anti-glioma agent (interleukin-13 bound to Pseudomonas toxin, IL13-PE) to the brainstem while monitoring drug distribution with a co-infused surrogate MR-imaging tracer (gadolinium-DTPA). Based on the safe and successful use of this delivery model in rodents and primates, we developed a clinical protocol to treat diffuse brainstem gliomas in pediatric patients with IL13-PE co-infused with gadolinium-DTPA. We safely treated 5 patients with CED of IL13-PE and gadolinium-DTPA and successfully tracked the distribution of drug in real-time using intraoperative MR-imaging. We published an analysis of the accuracy of direct magnetic resonance imaging-guided placement of drug infusion cannulae in study subjects. We submitted for publication our report of IL13-PE drug distribution and clinical effects and have responded to questions from the reviewers. Our early findings provide foundational data on monitoring drug delivery and intratumoral treatment of diffuse brainstem gliomas, which may be applied to the treatment of other CNS malignancies including malignant gliomas. Neurodegenerative disorders. The properties of CED allow it to selectively manipulate distinct subsets of neurons (and other cell types) for therapy. We are investigating in a clinical trial a targeted gene-therapy approach to deliver the neurotrophic protein, GDNF, to the putamen in patients with Parkinson disease. In this condition, convection is being explored to selectively distribute AAV2-GNDF (adenoassociated virus type 2, carrying the human GDNF gene) and maintain dopaminergic neurons that would otherwise degenerate. We completed treatment of cohorts of 6 patients at 2 dose levels in this Phase I, dose-escalation study and treated 1 subject at the third dose level. The method provides a targeted, site-specific means of restorative neurosurgery. In laboratory animals, we completed a study of the effect of convection-enhanced delivery of muscimol, a GABA-A agonist. A solution of muscimol and gadolinium-DTPA was infused bilaterally into the subthalamic nuclei. Distribution of muscimol was monitored in real-time by observing the distribution of gadolinium-DTPA in the infusion solution. Behavioral changes, safety, and distribution of muscimol were recorded. A report analyzing drug distribution and behavioral effects was published this year. This work was performed to support a clinical trial of infusion of muscimol into the subthalamic nucleus during deep brain stimulation (DBS) surgery. This clinical study would provide insight into the potential mechanism of action of electrical stimulation of the subthalamic nucleus. This work could ultimately lead to chemical neurosurgery, in which patients with degenerative disorders could be treated using convection-enhanced delivery of agents acting on specific neurotransmitters and brain structures. Epilepsy. The hippocampus is the usual site of origin of medically intractable epilepsy. Relief of this type of epilepsy could occur if a method were developed to selectively suppress the epileptic focus within the hippocampus. After success in ablating seizures in a rodent model using convective perfusion of the epileptic focus, our laboratory conducted a study of the toxicity and distribution of the chronic infusion of muscimol into the hippocampus of 10 non-human primates. Depth electrode studies showed that electrical activity in the hippocampus could be suppressed by muscimol. Autoradiography of infused muscimol demonstrated that muscimol could be delivered to the entire hippocampus using convective perfusion. The infusions were tolerated without brain injury or permanent adverse effects. The FDA granted us approval for intracerebral CED of muscimol to brain. Candidates for seizure surgery were recruited for the clinical study of the infusion of muscimol into the hippocampus to temporarily inactivate the neurons of the epileptic focus. The first 3 of 18 subjects entered this trial and underwent 1 to 2-day infusions into the seizure focus of the study drug, muscimol (a GABA agonist) under an FDA IND. The infusions were well-tolerated, but recruitment of more subjects was not successful because short-term muscimol infusion did not offer permanent treatment of epilepsy. A manuscript describing the results of the study was accepted for publication this year. Based on this experience, we propose translational development of other agents for medically-intractable epilepsy that will permanently and selectively inactivate the epileptic focus.